I have a custom designed and built CoreXY printer. I have noticed that when I order the X to move from one side to the other (right to left) that the Y axis moves back about 0.6 mm as well. When I move the X back (from left to right) the Y also moves forward approximately the same amount.

What property of the CoreXY system could be causing this? Note, I am using 608 bearings instead of toothed idler pulleys; which I hope to rectify soon.

Steps per mm from the Marlin configuration.h:

DEFAULT_AXIS_STEPS_PER_UNIT   { 475.79, 482.87, 468.66, 188.1 }

Note, they are not the same for X and Y. This is because I am trying to run a calibration cube test on the machine.

  • 1
    $\begingroup$ that sounds like either a firmware or a hardware problem... $\endgroup$ – Trish Oct 25 '19 at 21:26
  • $\begingroup$ Is the motor turning causing the movement? Or is this a measurement you've taken for the y-axis travel? $\endgroup$ – Pᴀᴜʟsᴛᴇʀ2 Oct 25 '19 at 22:50
  • $\begingroup$ @Paulster2, This is a measurement I've taken for the y-xis travel; $\endgroup$ – user77232 Oct 26 '19 at 12:03
  • 1
    $\begingroup$ Are the steps per mm for X and Y set to the same value? $\endgroup$ – Tom van der Zanden Oct 26 '19 at 12:12
  • 1
    $\begingroup$ @user77232 it could be a pinning error (as in the firmware pins are messed up) or steps/mm (which is in the firmware). try to check hardware first. $\endgroup$ – Trish Oct 26 '19 at 12:24

For a CoreXY printer to move an axis (X or Y) it requires both stepper motors to turn. If both turn the same direction (at the same speed with the same pulleys), the X-axis will move, if they rotate both in a different direction (at the same speed with the same pulleys) the Y axis will move (see image of CoreXY kinematics below). Rotation of a single stepper would cause a 45° printing pattern.

CoreXY kinematics

All four corner points (two top pulleys and two bottom steppers) are fixed to the CoreXY frame

It is therefore highly illogical that you need to use different values for your steps per mm (X = 475.79 steps/mm, Y = 482.87 steps/mm). If you need to use such values, this implies that your mechanical layout/mechanics is/are incorrect (skew frame, different pulley diameters, slip on pulley, incorrect tension of belts, etc.). It is advised to make them equal and check the mechanical layout and inspect all parts (pulley diameters) and make the tension equal (e.g. using a Belt Tension Gauge). If the head still moves in the perpendicular direction as commanded, you could try to calibrate from there.

  • $\begingroup$ ok I assert that it is possible that there are variances in the tolerance of the motors, pulleys and belts. The drift in the values was my attempt to compensate for those variances. Would that be something that I compensate for in hardware or firmware? $\endgroup$ – user77232 Oct 29 '19 at 0:34
  • $\begingroup$ @user77232 I doubt that the manufacturing tolerances of steppers, pulleys and belts would make that much of an impact. My CoreXY is running with equal X and Y steps per mm and produces square products. $\endgroup$ – 0scar Oct 29 '19 at 13:17
  • $\begingroup$ Ok, so I need an actual answer: Do I reset the steps/mm in the firmware and try invest in sorting out the hardware or do I mess with marlin to get it to straighten out? $\endgroup$ – user77232 Oct 29 '19 at 13:42
  • 1
    $\begingroup$ @user77232 set them to be equal first, then calibrate. $\endgroup$ – Trish Oct 29 '19 at 14:09
  • 1
    $\begingroup$ @user77232 Try getting the tension in the belts to the same using a tension gauge, I've updated my answer. This gauge helped me too. $\endgroup$ – 0scar Oct 29 '19 at 15:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.